Evaluating and Refining a Pain Quality Information Visualization Tool with Patients and Interpreters to Facilitate Pain Assessment in Primary Care Settings

Abstract

Information visualization (InfoViz) tools offer a potential solution to pain communication challenges. Incongruencies in communication styles between patients with limited English proficiency (LEP), interpreters, and providers contribute to significant disparities in pain care and outcomes. This study’s purpose is to evaluate and refine a culturally appropriate InfoViz pain quality assessment tool for LEP Hmong patients. We conducted a three-part iterative user-centered study with LEP Hmong, bilingual Hmong, and Hmong interpreters with (1) participatory design sessions to evaluate and refine pain infographics for inclusion on the tool, (2) card-sorting to organize the infographics to match the mental models of LEP patients, and (3) a tool assessment to identify which tool accurately represented LEP patients’ mental models and was preferred in clinical settings. Fifty-five participants provided three common themes for pain infographics refinement: culturally-relevant colors, infographics resembling human anatomy, and action-specific squiggle lines. The card-sorting sessions revealed three organizational themes: sensation (n = 15; 71.4%), localization (n = 6; 28.6%), and severity of pain quality (n = 5; 24.3%). Most participants selected the localization as the most accurate tool and preferred it in clinical settings. Using a multi-step, user-centered approach resulted in a culturally appropriate pain InfoViz tool for LEP Hmong patients.

Introduction

Information visualization (InfoViz) tools, defined as visual representations of information (Krum 2004), offer a potential solution to communication challenges in healthcare settings. These tools can help patients and providers understand, discuss, and decide on treatment for health problems, including pain (Lor et al. 2018; Shuwandy et al. 2016; Bhavnani et al. 2016). To identify causes and effectively treat pain, information about the pain quality, location, and severity are vital. However, patients often express pain information (i.e., pain quality) in metaphorical or culturally based terms that do not match the medical terms of the providers (Lor, Rabago et al. 2020), resulting in suboptimal patient-provider communication.

The limited English proficiency (LEP) of patients amplifies the communication challenge. Thus, providers and LEP patients may rely on medical interpreters to communicate. Nevertheless, the interpretations may be inaccurate or misleading due to a lack of interpreter fluency or generational differences in language use (Lor et al. 2014). Hence, an effective InfoViz tool for pain communication has the potential to overcome translation challenges, allowing patients to select infographics that match their subjective experiences of pain. In this study, we report on the multi-step, user-centered design of an InfoViz tool for pain quality assessment in one LEP patient group: the Hmong.

Hmong Pain Communication

The Hmong in the U.S. are an ethnic group from a primarily agrarian society in Laos, arriving as refugees between the mid-1970s and 2000s (Pfeifer et al. 2012). There are 327,000 Hmong living in the U.S. (Budiman 2021), with 90% having LEP (Duffy 2007). The Hmong traditionally communicate verbally, having an oral tradition (Duffy et al. 2004). A written language, the Romanized Popular Alphabet, was developed by Christian missionaries in the 1950s but many Hmong are unfamiliar with it. However, most of the younger Hmong population are proficient in English and are bicultural (Hmong National Development Inc 2013).

In the Hmong language, pain is understood and communicated through cultural metaphors (Lor, Rabago et al. 2020; Lor et al. 2014; Lor and Chewning 2016). For example, Hmong patients describe throbbing pain as “it hurts like chicken pecking” (Lor, Rabago et al. 2020), reflecting their agrarian roots (Lor 2018; Lor et al. 2016; Lor et al. 2013; Xiong 2018) not captured in American medical terminologies. Thus, Hmong pain language does not conform to the language of pain used by providers (Lor, Vang et al. 2020). Hence, interpreters have struggled to interpret the metaphors (Lor et al. 2022), placing the Hmong at exceptionally high risk of pain-communication misunderstandings and disparities in pain care. Acknowledging the communication difficulty, Hmong patients and healthcare providers desire to have a pain assessment InfoViz tool to facilitate pain communication (Lor, Vang et al. 2020).

Pain Quality InfoViz Tools

While InfoViz tools have been used in pain medicine to communicate about pain for decades, they have focused on pain severity (Wong and Baker 2001; Choinière and Amsel 1996; Hicks et al. 2001), created primarily for children (Wong and Baker 2001). There is limited focus on assessing pain quality using InfoViz tools (Lalloo and Henry 2011; McAuley 2006; Padfield et al. 2010; Rossato and Magaldi 2006). The four tools located use different InfoViz approaches, including pictures (McAuley 2006), icons (Lalloo and Henry 2011), photographic montage (Padfield et al. 2010), and visual cards (Rossato and Magaldi 2006), deemed easy to use and feasible. However, they have not been developed for and tested on LEP adult patients. Since 80% of U.S. hospitals encounter LEP patients (Kang et al. 2010), it is vital to expand the InfoViz landscape for pain assessment (Kang et al. 2010).

Study Purpose

This study aims to evaluate and revise a culturally appropriate pain quality assessment InfoViz tool through a user-centered design with LEP Hmong and medical interpreter end users. The goal is to develop a culturally appropriate tool to facilitate pain communication among LEP Hmong patients, interpreters, and providers. The study comprises three parts: 1) participatory design sessions with LEP and bilingual Hmong participants and interpreters to evaluate and refine the pain infographics included in the tool, 2) card sorting with LEP and bilingual Hmong participants and interpreters to organize the tool’s pain quality infographics to match the mental models of LEP patients, and 3) a tool assessment to identify which tool accurately represents LEP patients’ mental models and is preferred for use in clinical settings.

Methods Overview

Our university’s institutional review board approved this study. Given the iterative nature of the three study parts, the methods and results are presented in sequence with a combined discussion of all parts. At the end of each session, each participant received $50. Table 1 contains a summary of all three studies.

Table 1.

Summary of Purpose, Procedures, and Components for Each Study Part

	Part 1a: Infographic Refinement	Part 1b: Infographic Evaluation	Part 2: Card Sorting	Part 3: Tool Development and Evaluation
Purpose	To evaluate and provide suggestions for refining an initial set of 54 pain quality infographics	To evaluate and provide suggestions for refining the reduced set of pain quality infographics from Part 1	To organize the 16 pain quality infographics to match the mental models of LEP Hmong adults	To identify which pain quality InfoViz tool accurately represents LEP patients’ mental model, facilitates better communication, and is the most preferred in a clinical setting
Participants	LEP Hmong adults (n = 27) and bilingual Hmong adults (n = 15)	Family and professional Hmong interpreters (n = 13)	LEP Hmong adults (n = 6) and bilingual Hmong adults (n = 15)	*Bilingual young adults (n = 10) and Hmong interpreters (n = 13)
Tasks	(1) Share perceptions of each infographic’s meaning (2) Select the infographic that best represents the intended meaning of the pain quality (3) Provide feedback on revising the infographic to improve accuracy, comprehension, and cultural appropriateness	(1) Review the PDF pain quality infographic designs document one week before session (2) Provide feedback on revising the infographic to improve accuracy, comprehension, and cultural appropriateness	(1) Individually sort the pain quality cards into piles that make sense to the participant and categorize them (2) Share aloud the sorted piles and categories (3) As a group, vote on one set of sorted piles (4) As a group, rearrange the cards into categories that match the participants’ mental models	(1) Reflect on experiences with LEP Hmong adults in clinical settings (2) Select the tool that best corresponds to LEP Hmong patients’ mental models
Total Sessions	8	8	6	11
Data Collection Setting	Four in-person sessions and four virtual sessions	All virtual sessions	Three in-person sessions and three virtual sessions	All virtual sessions
Use of Diagram	Pre-session: PDF document During session: Microsoft PowerPoint with pain quality infographic	Pre-session: PDF document During session: Microsoft PowerPoint with pain quality infographic	In-person session: set of sixteen 5” x 3” notecards containing a single pain quality infographic each and twenty-two 5” x 3” notecards depicting various foods and drinks Pre-virtual session: Google Jamboard example video, and Google Jamboard card-sorting activity link During virtual session: screenshots of completed Google Jam card-sorting activity and categories via PowerPoint	During virtual session: Microsoft PowerPoint of the three InfoViz tools

Note:

^*Two interpreters did not participate in the tool and text preference due to poor internet connection during the WebEx virtual meeting.

Part 1: Participatory Design Sessions With Limited English Proficiency and Bilingual Hmong Adults and Interpreters

Materials and Methods

This participatory design study was conducted in the U.S. Midwest from October 2019 to July 2021. The purpose was to evaluate and refine an initial set of infographics representing and communicating Hmong patients’ pain experiences.

Participants

We recruited a purposeful sample of LEP Hmong adults, bilingual Hmong adults, and Hmong interpreters from one large metropolitan city and one small town. We chose to include bilingual Hmong participants and interpreters in this study to obtain additional perspectives from individuals who speak both Hmong and English and are the end users of the tool. Moreover, they provided content validity to assist with the comprehension and refinement of the infographics. We posted flyers at a Hmong organization that serves over 1,000 individuals in the metropolitan city to recruit patients and used word of mouth to recruit interpreters. We also contacted a Hmong home health nurse in the small town to share our flyers and used word of mouth to recruit LEP and bilingual Hmong adults.

To be eligible, both Hmong LEP and bilingual adults had to self-identify as Hmong, be 18 years or older, and have self-reported pain at the time of recruitment. We identified LEP participants using the U.S. Census in the American Community Survey question, “How well do you speak English?” as a proxy for LEP (United States Census Bureau 2020). Bilingual Hmong adults were included if they self-reported as proficient in Hmong and English. Interpreters, professional or lay, were eligible if they were 13 years or older and self-identified as an interpreter for Hmong individuals in a healthcare setting. Consistent with the Affordable Care Act, Section 1557, children of any age can act as medical interpreters for their family members, provided the adult patient consents (Squires and Youdelman 2019).

Data Collection

Part 1a: Infographic Design Refinement with Limited English Proficiency and Bilingual Adults

We held four virtual sessions from November 2020 to March 2021 using WebEx to accommodate for the coronavirus disease 2019 (COVID-19) pandemic and four in-person sessions from March 2021 to July 2021 at the Hmong organization or in participant homes following COVID-19 protocols. In Part 1a, LEP and bilingual Hmong adults evaluated and provided suggestions for refining an initial set of 54 pain quality infographics created in a previous study (Lor et al. 2020). We conducted 12 iterative design sessions (n = 8 LEP adults; n = 1 older bilingual adult; n = 3 younger bilingual adults) with two to six adults. We first conducted the sessions with LEP adults, followed by bilingual adults. Before the design sessions, we distributed 8.5” x 11” Portable Document Formats (PDFs) via email or ground mail containing a single pain quality infographic on each page and asked the adults to review the designs.

During the sessions, we obtained verbal consent and used Microsoft PowerPoint to present two to four infographics for each pain quality (e.g., shooting, sharp, dull, and achy) without information indicating which pain quality the infographic was intended to represent. We asked LEP and bilingual Hmong adults to share their perceptions of each infographic’s meaning. We then revealed the intended meaning of the infographic and asked the LEP and bilingual Hmong adults to select the infographic that best represented the intended meaning of the pain quality and provide feedback on how to revise the infographic to better match their perceptions of the specific pain quality metaphor. Designs that were not selected as an adequate representation of the metaphor or that the LEP and bilingual Hmong adults perceived as inaccurately reflecting the metaphor, challenging to understand, or not culturally appropriate were eliminated from future iterations.

Part 1b: Interpreter Evaluation of Infographics

Part 1b involved interpreter evaluations of the infographics over eight sessions with one to three interpreters per session. Due to the COVID-19 pandemic, we only held virtual sessions for interpreters from June to July 2021. All the sessions were scheduled at least one week apart to allow time for the research team to modify the infographics before the next session.

We emailed the interpreters a PDF document containing the pain quality infographic designs one week before the session. We asked them to evaluate the accuracy of the correspondence between the intended meaning of the infographics and the pain quality as well as the comprehensibility and cultural appropriateness of the infographics. All the design sessions were audio recorded. At the end of each session, the interpreters completed a demographic survey.

Analysis

A team of bilingual and bicultural student transcriptionists (BX, NY, and MCV) transcribed the interviews. The lead researcher (ML), who is bilingual and bicultural in Hmong and American communities, reviewed all the transcripts for accuracy. For any discrepancies, the lead researcher and student transcriptionists listened to the interviews together. Corrections were made based on consensus.

Using Excel, we summarized the participants’ demographic information with descriptive statistics to manage the design session data while applying summative and directed content analyses to the participatory design session data (Hsieh and Shannon 2005). These two analysis approaches were used to analyze the responses from the LEP and bilingual Hmong adults as well as the interpreters because they received the same questions, and their responses were similar across categories. Therefore, in the remainder of the methods and results sections, we refer to LEP and bilingual Hmong adults and interpreters as “participants.”

Two team members independently read the transcripts and initially coded them using predetermined categories: accuracy, comprehension, preferences, and feedback. Two coders relistened to the digitally recorded design sessions.

We defined accuracy according to whether the participants stated exactly or had terms consistent with Hmong pain quality metaphors, culturally based terms, or descriptors. For example, if participants stated, “The chicken image represents hurting like a chicken pecking,” it was coded as “accurate.” Responses deviating from the metaphors, culturally based terms, or descriptors were coded as inaccurate. Infographics that participants inaccurately identified as representing the intended pain metaphor were not included in subsequent design sessions.

We defined comprehension according to whether the individual understood the presented infographic. To assess participants’ understanding, we listened for cues indicating confusion. For example, if a participant said, “I am not sure what the visual means,” it was coded as “lack of comprehension.” Design saturation was achieved when new ideas for improving the infographics ceased to be generated and no confusing design elements were identified.

Findings

Table 2 contains all the demographic characteristics of the LEP and bilingual Hmong adults and interpreter participants.

Table 2.

Participant Social Characteristics for study Parts 1 and 2 (n=55)

	LEP Hmong (n=27*)		Bilingual Hmong (n=15*)		Interpreter (n=13)
	Count	Percent	Count	Percent	Count	Percent
Gender
Male	9	33.3	5	33.3	2	15.4
Female	17	63	9	60	11	84.6
Age
Mean	62.2	-	39.5	-	38.6	-
Standard Deviation (SD)	7.7	-	18.6	-	8.1	-
Length of Stay in the US
Mean	35.2	-	24.2	-	34.3	-
SD	6.0	-	9.8	-	6.4
Location of Birth
United States	-	-	-	-	4	30.8
Laos	-	-	-	-	5	38.5
Thailand	-	-	-	-	4	30.8
List of Pain
None	5	18.5	6	40	-	-
Upper Extremities: Shoulder, Hand, Arm	5	18.5	2	13.3	-	-
Lower Extremities: Feet, Legs, Knees	11	40.7	2	13.3	-	-
Headaches	2	7.4	1	6.7	-	-
Neck	1	3.7			-	-
Back	10	37	4	26.7	-	-
Fatigue	1	3.7	-	-	-	-
Pain associated with Diabetes	1	3.7	-	-	-	-
Toothaches	1	3.7	-	-	-	-
Stomach Pain	1	3.7	-	-	-	-
Taking Medications for Pain
Yes	9	33.3	1	6.7	-	-
No	17	63	12	80	-	-
Health Insurance
Employer	3	11.1	6	40	-	-
Medicaid	20	74.1	6	40	-	-
Medicare	0	0	0	0	-	-
Medicare/Medicaid	0	0	1	6.7	-	-
None	3	11.1	0	0	-	-
Education
No Education	25	93	3	20
High school	1	3.7	1	6.7	1	7.7
Some College	-	-	2	13.3	2	15.4
Associates	-	-			4	30.8
Bachelors	-	-	3	20	5	38.5
Masters	-	-	3	20	1	7.7
Professional	-	-	1	6.7
Type of Interpreter a
Staff Interpreter	-	-	-	-	7	53.8
Certified Interpreter	-	-	-	-	1	7.7
Contract Interpreter	-	-	-	-	4	30.8
Family Interpreter	-	-	-	-	9	69.2
Years of Interpreting
Mean					15
SD	-	-	-	-	8.4
Literacy Level
Write in Hmong
Very well	3	11.1	3	20	2	14.3
Pretty well	9	33.3	3	20	1	7.1
Not too well	5	18.5	7	46.7	7	50.0
Not at all	9	33.3	1	6.7	3	21.4
Read in Hmong
Very well	3	11.1	4	26.7	2	14.3
Pretty well	12	44.4	4	26.7	5	35.7
Not too well	5	18.5	6	40	4	28.6
Not at all	6	22.2	0	0	2	14.3
Write in English
Very well	0	0	7	46.7	8	57.1
Pretty well	0	0	5	33.3	5	35.7
Not too well	13	48.1	2	13.3	-	-
Not at all	13	48.1	0	0	-	-
Read in English
Very well	0	0	8	53.3	10	71.4
Pretty well	0	0	5	33.3	3	21.4
Not too well	18	66.7	1	6.7	-	-
Not at all	8	29.6	0	0	-	-

Note:

^*missing data from 1 participant.

^aThe total number in this category may be more than 14 because interpreters can indicate to have more than one position such as they can be a professional and family interpreter.

Design Suggestions and Implementation

Several common themes, described in this section, arose from participants’ suggestions during the participatory design sessions. Moreover, participants suggested adding twisting as another pain quality to the list of infographics.

Design saturation—the point at which no new ideas for improving the design were generated—was achieved for internal and external burning, cramping, pinching, pins and needles, pressure, sharp, squeezing, throbbing, and twisting by the first three design sessions. Reaching design saturation for the other pain qualities required four sessions for aching and numbness, five sessions for sore, six sessions for tingling, and eight sessions for shooting and dull. Figure 1 shows the evolution of the infographic design for a specific pain quality, dull, throughout the sessions.

Figure 1.

Exemplar of Design Evolution of Dull Infographic

Use of Culturally Relevant Colors

Throughout the design sessions, participants consistently emphasized two colors that have cultural meanings as visual symbols: red and white. Culturally, red indicates pain or hurting, and different shades indicate the pain severity: light red conveys less intense pain, while dark red is greater or severe pain. Hence, participants suggested adding red to pain qualities they perceived as painful. One participant suggested that we “add red to the thighs, calves, back, arms, and hands so that we know that it is sore” (Participant 22). Accordingly, we added red to the aching, shooting, tingling, pinching, and sore infographics.

Several participants suggested that creamy white should represent numbness to distinguish it from pain since they did not associate numbness with pain. One participant explained, “I don’t think of red though as numbness. … When I think of numbness, I think more of a pale color” (Participant 43). Consequently, we added white to the numbness infographic.

Infographics Resembling Human Anatomy

Participants wanted “a human resemblance” embedded in the infographics to elucidate each pain quality. Many expressed that without a human figure, it was unclear whether the pain was “real,” and they could not understand what the pain was, particularly for dull, cramping, tingling, numbing, sore, aching, shooting, pressure, pinching, and sharp pain qualities. Participants shared that these pain qualities often occur in a specific location, so it would aid their understanding of the infographics if the infographics were consistent with their experiences. For example, concerning shooting pain, one participant requested including a body on the infographic, focusing on the shoulders, back, and arms: “If you don’t draw the person, you won’t know what this is. You need to draw the person so you know that the pain is … moving from the back” (Participant 15).

Another example was that because cramping can occur in the stomach or leg, the cramping infographic should include the leg. However, for the four pain qualities of throbbing, pins and needles, burning, and twisting, most of the participants expressed that they did not need the connection of a human body because the infographics were clear and accurate. Consequently, body locations were added to dull, cramping, tingling, numbing, sore, aching, shooting, pressure, pinching, and sharp pain infographics.

Action-Specific Squiggle Lines

Participants consistently requested that squiggle lines be added, along with action-specific events, to indicate cramps, numbness, soreness, pressure, and achiness. For many of the participants, squiggle lines represented the sensation they experienced with the stated pain qualities. For example, many participants suggested representing tingling by adding a squiggle line to the leg of a human body to illustrate the sensation of something crawling on the leg. Hence, we added squiggle lines to the infographics for cramps, numbness, soreness, pressure, and achiness.

Accuracy and Comprehensibility of Infographics

The LEP and bilingual Hmong participants consistently rated eight of the pain quality infographics—burning, cramping, pinching, pins and needles, pressure, sharp, throbbing, and twisting—as accurate and comprehensible throughout the eight design sessions. In contrast, the participants poorly comprehended and variously and inaccurately interpretated the intended meanings of six of the pain quality infographics—aching, dull, numbness, tingling, shooting, and sore. Except for numbness and tingling, the variations in interpretation and poor comprehension were due to a lack of direct translations or comparable metaphors in the Hmong language and cultural constructs. One participant elaborated, “Us older people only know that pain is pain, then we don’t know how to further describe it.” (Participant 4).

Therefore, all participants, including LEP and bilingual Hmong adults and interpreters, recommended that researchers couple the infographics with text. For example, one participant suggested, “Just write that it is dull” (Participant 4). An interpreter shared, “It helps me … interpret the explanation of the symptoms….It tells you what you’re describing to the patient” (Participant 40). Although the LEP participants could not read the text, they said that an interpreter or a family member could read it to them.

Part 2: Card Sorting with Limited English Proficiency and Bilingual Hmong Adults and Interpreters

Materials and Methods

Following the participatory design sessions described in Part 1, we narrowed the number of pain infographics to 16. We conducted card-sorting sessions to gather data on organizing the 16 pain quality infographics on the InfoViz tool according to the mental models of the LEP Hmong adults (Stone et al. 2005; Brinck et al. 2001; Courage and Baxter 2005). These sessions were held in the same metropolitan city and small town as in Part 1 from April 2021 through May 2021.

Participants

We recruited six LEP participants from the previous design sessions and 15 new bilingual participants for the study (N = 21). The same eligibility criteria and recruitment methods were used as in Part 1 of the study.

Data Collection

We conducted six sessions with one to five adults in each session (n = 6 LEP adults; n = 5 bilingual older adults; n = 10 bilingual younger adults). The rationale for including bilingual Hmong was to determine whether they and the LEP participants shared similar mental models and to gain a variety of organizational styles for the tool. Two LEP and one of four bilingual sessions were conducted in person in adherence with COVID-19 protocols at the Hmong organization. The remaining sessions were conducted virtually via WebEx.

In-Person Sessions

Using the finalized pain quality infographics from Phase 1, we created sets of 16 5” x 3” notecards containing a single pain quality infographic each. Because participants were new to card sorting, we also created a set of 22 5” x 3” notecards depicting various foods and drinks, Hmong and American desserts, meats, and fruits to demonstrate the card-sorting exercise. We gave participants both sets of notecards at the beginning of the sessions and then demonstrated the process of thinking, putting the notecards in piles, and labeling them. We provided two examples of organizing the piles: separating them into healthy versus non-healthy foods and categorizing them according to food categories such as fruits, vegetables, meats, and drinks. We encouraged participants to use the food and drink cards until they felt comfortable with the task before progressing to the pain cards.

Each participant sorted the pain quality cards into piles that made sense to them and categorized them. They also completed a survey about their experience. We took photographs of the sorted piles and recorded the participants’ explanations of their thought processes.

Thereafter, all the session participants gathered to build a consensus on the sorted categories. This task began with each person sharing aloud their categories and piles. After the explanations, participants voted on one set of sorted pain quality cards that made the most sense to them. Collectively, they rearranged the highest-voted cards into categories that matched their mental models.

Virtual Sessions

The virtual sessions were conducted with bilingual Hmong. Because the sorting activity was virtual, we created a video using the same content from the in-person sessions and demonstrated the card-sorting activity via Google Jamboard, a digital interactive whiteboard. The Google Jamboard example video, and link to the Google Jamboard card-sorting activity were sent to participants via email one week before the virtual session. Participants completed the card-sorting activity before the session.

We took screenshots of the completed Google Jamboard sorting and categories and shared the images during the session. As with the in-person sessions, the participants first shared their categories and piles aloud and subsequently voted on the set of cards that made sense to them and rearranged the highest-voted cards into categories that matched their mental models.

Analysis

Because all participants created unique card-sorting categories, each category was entered into an Excel spreadsheet. We applied a thematic analysis to the data because there were overlaps in categories. For example, we grouped participant categories such as “extreme pain,” “moderate pain,” or “low pain,” under the theme “severity of pain quality.”

Findings

Card-Sorting Themes

We identified three main themes among the participant centered-created categories: 1) sensation (n = 15; 71.4%), 2) localization (n = 6; 28.6%), and 3) severity (n = 5; 24.3%¹; see Table 3). The sensation theme contained pain qualities that participants defined as physical feelings, such as throbbing, tightness, burning, aching, dullness, soreness, and sharpness. The localization theme involved pain qualities that were easily pinpointed or located in a generalized area. Examples of easily pinpointed pain qualities included pins and needles, pinching, dull, and aching, whereas more generalized pain qualities included tingling, numbness, soreness, and pressure. The severity theme described different levels of pain qualities. For example, lower pain severity included pins and needles, pinching, throbbing, and dull, while greater pain severity included tingling, numbness, soreness, and shooting.

Table 3.

Results of The Three Sorted InfoViz Tools and Interpreter Tool Preferences

Part 2: Results of The Three Sorted InfoViz Tools (N = 21*)
	Tool 1: Sensation	Tool 2: Severity	Tool 3: Localization
Description	Organized the first tool by two types of pain sensation: depth (first three columns) and tightness (last column), reading from top to bottom for each column.	Organized the second tool by pain severity, reading from left to right and beginning with minimal severity at the uppermost rows and progressing to the most severe pain at the bottom rows.	Organized based on localized (first two rows) versus generalized or widespread pain (last two rows), reading from left to right in rows
Part 3: Results of Interpreter Tool Preferences (N = 23*)
Young Adult Bilingual (n=9)	5	0	4
Interpreter (n=12)	3	0	9
Total	8	0	13

Note:

^*Two interpreters did not participate in the tool and text preference due to poor internet connection during the WebEx virtual meeting.

Part 3: Tool Development and Evaluation

Tool Development

We consulted with a survey methodologist on the themes developed from Part 2 and used the common themes to organize the pain infographics into three potential tools (Table 3). We organized the first tool by two types of pain sensation: depth (first three columns) and tightness (last column), reading from top to bottom for each column. For example, the depth infographics reads from the top column with superficial to the bottom with deep into the skin. We organized the second tool by pain severity, reading from left to right and beginning with minimal severity at the uppermost rows and progressing to the most severe pain at the bottom rows. We organized the third tool based on localized (first two rows) versus generalized or widespread pain (last two rows), reading from left to right in rows. After building the tools, we brought them to new bilingual young adults and interpreters to gather their preferences and feedback. The tools were built using Microsoft PowerPoint. Each tool comprised 16 pain quality infographics organized based on the card-sorting theme and displayed on a PowerPoint slide for the participants to review.

Methods and Materials

Sample and Setting

We recruited 10 bilingual young adults and 13 Hmong interpreters, all new to the study, from the same locations and using the same inclusion criteria as in Parts 1 and 2. Bilingual adults and interpreters were used instead of LEP adults to provide content validity. Because they spoke both English and Hmong, the participants were able to provide feedback on the linguistic and cultural appropriateness of the tool.

This tool is intended to facilitate patient-provider communication in scenarios with a bilingual and bicultural interpreter. Hence, having bilingual and bicultural individuals evaluate the tools was necessary. Eleven sessions were held from May 2021 to July 2021.

Data Collection

The tool selection sessions aimed to identify which pain quality InfoViz tool accurately represented LEP patients’ mental model, facilitated better communication, and was the most preferred in a clinical setting. We presented all three tools that the LEP adults designed and asked the bilingual young adults and interpreters to think of experiences they had with LEP Hmong adults in a clinical setting. For example, we instructed, “Think back to when you worked with LEP Hmong adults. How did they present their pain information? As you view these tools, keep that framework in mind and select the tool you think best corresponds to LEP Hmong patients’ mental models.” Each session included one to five participants and was conducted virtually via WebEx.

Bilingual Sessions

We showed the bilingual young adults three pain quality InfoViz tools (Table 3) and asked them several questions: “How do you understand this?” “Which organization of these three tools will help LEP Hmong best describe their pain?” “What do you like and not like about each tool?” “Should we include both Hmong and English translations?” and “How can we make this tool better?” We recorded all the sessions and made changes to the pain quality InfoViz tools before the next session.

Interpreter Sessions

We showed the interpreters the pain quality InfoViz tools (Table 3) and asked about their reactions to the tools, the English and Hmong text placement, and the types of training needed to use the tool. Because there was more than one Hmong pain metaphor phrase for each pain quality, we asked the interpreters about their preference regarding the number of metaphors and the translation texts for each infographic. For example, acknowledging that the interpreters may have varied skills in interpreting Hmong metaphors, we asked them if they desired a direct translation from Hmong to English. We also requested that they provide feedback on the elements of the tools, including checkboxes, shapes, and placement of the pain quality visualizations and text. For example, we asked whether it would be helpful if patients were instructed to circle pain qualities instead of using checkboxes. We recorded all the sessions and made changes to the tools before each subsequent session.

Findings

Twenty-three participants (n = 13 interpreters; n = 10 bilingual adults) undertook this tool assessment. Because the responses from the bilingual adults and interpreters were similar, we report them below together.

Tool Preferences

Thirteen of the 23 participants (n = 13 [n = 4 bilingual adults; n = 9 interpreters²]) preferred the tool to be organized based on centralized versus widespread pain categories. Eight of the 23 participants (n = 8 [n = 5 bilingual adults; n = 3 interpreters]) preferred the tool to be organized according to sensation. No participant preferred the tool to be organized by the severity of pain quality (Table 3).

Text Placement Preference

All the bilingual adults and 11 of 13 interpreters preferred to have the English text placed above the infographic with the Hmong text below. The most frequent reason for this preference was to convey the response directly to the provider. For example, one interpreter stated, “If you think about the interpreter, they interpret in English. … When you write ‘mob mob tag’ [it hurts everywhere], you’re still going to have to interpret that in English” (Participant 44).

Some bilingual adults wanted only the Hmong text included to assist literate Hmong by anchoring the Hmong phrase with the infographics. Likewise, most of the interpreters wanted to retain the Hmong text, acknowledging that some interpreters cannot effectively interpret Hmong passages in English. One interpreter explained, “Because the Hmong language does not always have the words to directly translate to English, … we would have to describe the definition they want [in Hmong]” (Participant 45).

Only one interpreter disagreed and preferred to have the Hmong text above and the English below. This participant reasoned that because the tool is for Hmong patients, it is important to prioritize the Hmong text. Thus, literate Hmong patients would see the Hmong phrase first, facilitating their understanding of the pain visualization.

Quantity of Hmong Descriptor Text

Most of the interpreters (n = 8; 61.5%) preferred including the most common phrases they had heard among Hmong patients instead of all possible Hmong metaphors or descriptors for each pain quality. The reasons for only using the most common phrases included to avoid overloading and distracting from the text, overpowering the infographic’s purpose, or requiring too much time to read. One interpreter explained, “I think that it’s too much. I think that we should include the most common phrase people use because many Hmong people don’t use all the Hmong phrases too.” Another interpreter asserted, “If you use all the phrases for one picture, then it will get lost.” (Participant 47). Only one interpreter suggested having at least two of the most common phrases as they had observed patients describing pain qualities differently:

The most common is the right way, but some of them don’t use it like that. They don’t use the vocabularies that we use, so they won’t know it. To cover our bases … we should put two phrases, even if it’s a little bit more. (Participant 41)

In contrast, three of the interpreters (23%) preferred using every Hmong phrase because older LEP Hmong patients describe their pain differently. Hence, using every pain phrase would best capture all the terms used to describe the pain quality: “Language doesn’t always line up. … I think it’s better to have more instead of trying to concrete associate a term in Hmong with a term in English” (Participant 50).

Text Translation

Some of the interpreters (n = 4; 33.3%) preferred to retain the English translations because they wanted to better understand and interpret the pain qualities. One interpreter expressed that:

It’s helpful to have … so people are on the same page. In case, by chance, you didn’t know … that “tu leeg” refers to leg cramping, specifically, but also for the provider who is not a linguist and doesn’t realize that there’s not a perfect, direct correlation between every word in both languages. I think it will help them to see exactly what they’re talking about. (Participant 50)

However, half of the interpreters (n = 6; 50%) preferred not to have direct English translations because they could be redundant and visually busy. Moreover, they felt that interpreters should know how to describe the pain to the healthcare providers.

Elements of Tool Feedback

Most of the interpreters emphasized reducing the amount of text within the infographic and making it larger. One interpreter stated, “Many Hmong people don’t know how to read Hmong, and they don’t know English. The picture can reveal more. So, I think that if the picture was larger, then they would understand and see the picture better” (Participant 45).

All the bilingual adults and most of the interpreters preferred to have checkboxes to select the pain quality because it was easier for them to follow the directions. Additionally, one interpreter suggested, “You should include an example of a check in the checkbox. When Hmong people see this … if they cannot read, then they know that ‘Oh, we should checkmark that’” (Participant 45). Over half of the interpreters (n = 7; 53.8%) commented on the tool’s usability, and all agreed that it was understandable and feasible for use in a healthcare setting.

Discussion

This study conducted a multi-step, user-centered approach to iterate and evaluate an infographic assessment tool for pain quality to assist LEP Hmong patients and their interpreters communicate with healthcare providers. The study provided two new contributions to infographic and pain literature: 1) an evaluation of pain quality infographics with an LEP population and medical interpreters and 2) an incorporation of card sorting to ensure that the assessment tool matches the users’ mental models. Patient mental models were not included in the existing pain infographics (Stone et al. 2005; Brinck et al. 2001; Courage and Baxter 2005). This research fills a gap in healthcare pain infographics regarding a lack of such tools developed or tested with LEP users.

Consistent with prior research conducted with the Hmong population, red was a significant indicator of pain or something negative in this study (Lor et al. 2021). We also learned that the Hmong participants preferred an action-oriented and human-centered design for specific pain qualities that appeared to be more challenging to express in their culture. This finding is unsurprising as research supports the importance of designing infographic materials for familiar, simple, and engaging communication with personality (Baxter et al. 2021; Agency for Healthcare Research and Quality [date unknown]; Franconeri et al. 2021). Thus, we recommend that the final tool incorporate red and have an action-oriented and human-centered design for each pain quality.

More than 50% of our LEP Hmong adult sample were non-literate in Hmong, and over 96% had low literacy in English and Hmong, indicating that they had low health literacy. While having low health literacy implies difficulty interpreting health-related information, we do not think that our sample was impeded by this challenge because the LEP Hmong adults were not required to read or write; rather, they were asked to reflect on the Hmong pain phrases (i.e., metaphors they were familiar with) read out loud by the researcher, an approach consistent with Hmong oral culture (Duffy et al. 2004) that mitigates literacy challenges.

Our study went beyond developing tools to directly translate infographics from English to another language, an approach that has been common in patient education for decades in both clinical (e.g., the Mayo Clinic’s back pain infographic [Mayo Clinic, date unknown]) and public health settings (e.g., the Center for Disease Control and Prevention’s arthritis infographics [CDC, 2019]) for multiple cultures. We designed pain infographics that went beyond a direct translation and were instead based on the pain expression descriptions (e.g., metaphors) of the Hmong language. Our finding that more than just literal translation is necessary for infographic comprehension and accuracy highlights the need to consider differences in mental models and cultural norms when designing infographics.

This study’s methods and findings have implications for future InfoViz research. For instance, researchers could use participatory design and card sorting to develop communication tools for other LEP populations. Including the card-sorting activity added another novel source of feedback that helped refine our tool. Since LEP Hmong adults face health disparities due to language barriers, cultural factors, and experiences (e.g., refugees), developing culturally congruent infographics can prevent communication issues that negatively impact care (Espinoza and Derrington 2021; St.Amant 2015; Spinillo 2012). Moreover, our study demonstrates the importance of involving LEP and bilingual adults and interpreters in creating culturally congruent pain infographics. Thus, future researchers should identify LEP populations in healthcare organizations to serve and collaborate with them to develop communication tools.

The study involved several limitations. Because the pain metaphors, descriptors, and mental models of images were specific to the Hmong population, they may not be transferrable or generalizable to other LEP populations. However, the methods used in this study could be used for developing pain quality infographics with other groups. Due to the COVID-19 pandemic, we collected some data virtually. However, some of the participants’ engagement levels were impacted due to poor internet connections. Moreover, we did not test the tool’s psychometric properties or evaluate its usability and impact on pain communication among patients, interpreters, and providers in a clinical setting. Additionally, we did not assess whether the culturally appropriate pain representations matched providers’ understanding of pain associated with a specific disease. The next steps are to assess the tool’s psychometric properties; examine the feasibility of using the tool in clinical practice with LEP Hmong patients, interpreters, and providers; and investigate its impact on pain communication.

Conclusion

We used a multi-step, user-centered approach to iterate and evaluate pain quality infographics. We organized the infographics into three tools to match the mental models of LEP patients and selected the most accurate tool for use in a clinical setting. The outcome was the development of a culturally appropriate infographic assessment tool for pain quality in LEP Hmong patients. Hence, this study highlighted the importance of using multiple approaches to develop a culturally appropriate pain communication tool for LEP populations. We anticipate that the findings will increase the chances of a successful implementation in clinical settings.